Search

To utilize the full potential of semiconductor materials in device applications including solar cells, LEDs, and lasers, the ability to precisely and controllably tune the charge carrier concentration and hence the doping density is crucial. The conventional methods such as impurity doping with thermal diffusion or ion implantation, have been successfully implemented for doping bulk semiconductors for decades. In spite of the maturity of doping with traditional methods, it has remained a long-standing challenge to introduce impurity doping successfully into organic and new generation of semiconductors, such as conducting polymers and quantum dots. Additionally, the prospect of new technologies and the shrinkage in the device dimensions to nanoscale have stimulated researchers to search for alternative methods for achieving doping of such semiconductor materials reliably. Electrochemical doping is arguably the most powerful and versatile method for doping porous semiconductor materials, in which the charge carrier concentration can be precisely and controllably modulated as a function of applied potential by an external voltage source. Unfortunately, when the doped semiconductor film is disconnected from the voltage source, the electrochemically injected charges leave the film spontaneously in a matter of seconds to few minutes. In that regard, the stability of injected charges as well as the immobilization of external dopant ions need to fixed for achieving stable electrochemical doping of such semiconductor films to be used in device applications. The research carried out in this thesis is aimed to enhance the stability of injected charges and the fixation of dopant ions with photopolymerization treatment at room temperature in electrochemically doped quantum dots and conducting polymers. This was attempted by understanding the underlying mechanism of electrochemical doping in such porous films and eliminating or minimizing possible causes for in, ChemE/Opto-electronic Materials

Quantum dots (QDs) are considered for devices like light-emitting diodes (LEDs) and photodetectors as a result of their tunable optoelectronic properties. To utilize the full potential of QDs for optoelectronic applications, control over the charge carrier density is vital. However, controlled electronic doping of these materials has remained a long-standing challenge, thus slowing their integration into optoelectronic devices. Electrochemical doping offers a way to precisely and controllably tune the charge carrier concentration as a function of applied potential and thus the doping levels in QDs. However, the injected charges are typically not stable after disconnecting the external voltage source because of electrochemical side reactions with impurities or with the surfaces of the QDs. Here, we use photopolymerization to covalently bind polymerizable electrolyte ions to polymerizable solvent molecules after electrochemical charge injection. We discuss the importance of using polymerizable dopant ions as compared to nonpolymerizable conventional electrolyte ions such as LiClO4 when used in electrochemical doping. The results show that the stability of charge carriers in QD films can be enhanced by many orders of magnitude, from minutes to several weeks, after photochemical ion fixation. We anticipate that this novel way of stable doping of QDs will pave the way for new opportunities and potential uses in future QD electronic devices., ChemE/Opto-electronic Materials, Delft Energy Initiative, ChemE/Advanced Soft Matter

The increasingly complex nanoscale three-dimensional and multilayered structures utilized in nanoelectronic, catalytic, and energy conversion/storage devices necessitate novel substrate-selective material deposition approaches featuring bottom-up and self-aligned precision processing. Here, we demonstrate the area-selective atomic layer deposition (AS-ALD) of two noble metals, Pt and Pd, by using a plasma-polymerized fluorocarbon layer as growth inhibition surfaces. The contact angle, x-ray photoelectron spectroscopy (XPS), and scanning electron microscopy measurements were performed to investigate the blocking ability of polymerized fluorocarbon (CFx) layers against ALD-grown metal films. Both Pt and Pd showed significant nucleation delays on fluorocarbon surfaces. Self-aligned film deposition is confirmed using this strategy by growing Pt and Pd on the microscale lithographically patterned CFx/Si samples. CFx blocking layer degradation during ozone exposure was analyzed using XPS measurements, which confirmed the oxygen physisorption as the main responsible surface reaction with further hydroxyl group formation on the CFx surface. Our work reveals that the CFx layer is compatible with an ozone coreactant until the blocking polymer cannot withstand oxygen physisorption. Our results could potentially be used to investigate and develop radical-assisted AS-ALD processes for a wider selection of materials., ChemE/Opto-electronic Materials

No abstract available

The use of a hybrid fuzzy modeling can act as a good alternative in establishing a hydrocyclone control model in estimating the hydrocyclone parameter, d50c. In most control and engineering applications, the use of fuzzy system as a way to improve the human-computer interaction has becoming popular. The main advantage of this proposed hybrid fuzzy system used for hydrocyclone control is that it only presents a small amount of fuzzy rules. It uses memetic algorithms to optimize the fuzzy parameters of the system to yield in a more accurate hydrocyclone control system.

Fuzzy rule-based systems have been very popular in many engineering applications. In mineral engineering, fuzzy rules are normally constructed using some fuzzy rule extraction techniques to establish the determination model in predicting the d50c of hydrocyclones. However, when generating fuzzy rules from the available information, it may result in a sparse fuzzy rule base. The use of more than one input variable is also common in hydrocyclone data analysis. This paper examines the application of fuzzy interpolation to resolve the problems using sparse fuzzy rule bases, and to perform analysis of fuzzy interpolation in multidimensional input spaces.

In most control and engineering applications, the use of fuzzy system as a way to improve the human-computer interaction has becoming popular. This paper reports on the use of fuzzy system in mineral processing specifically in determining the parameter d50c of hydrocyclone. However, with the input-output data provided to build the fuzzy rule base, it normally results in a sparse fuzzy rule base. This paper examines the use of the improved multidimensional alpha-cut based (IMUL) fuzzy interpolation technique to improve the prediction capability of the sparse fuzzy rule base.

In electrical engineering, circuit theory is an essential subject normally delivered as lectures, tutorials and practical sessions. The cost to conduct practical sessions however is high due to smaller-size classes, the engagement of laboratory supervisors and the requirements on space, test equipment and consumable components. Although Computer Aided Learning (CAL) and simulator packages that claim to be interactive are used to replace practical experiments, the interaction is normally limited to the keyboard and the monitor. For first year students, a virtual laboratory environment can never replace the experience gained from hands on sessions. The objective of this project is to develop a "true-interactive" CAL package that functions as an electronic tutor, laboratory supervisor and assessor. The package consists of three parts. Part A introduces course topics with examples. Part B guides the student through several practical experiments. Students are expected to assemble and measure the electronic circuits as in a normal laboratory session. A custom-built hardware interface allows the software to monitor the student's progress. Finally, an assessment session will test the student's understanding on the subject. This paper provides the technical description of the development of package and discussions on possible further enhancements.

The accuracy in the estimation of hydrocyclone parameter, d50c, can substantially be improved by application of artificial neural networks (ANN). With ANN, many nonconventional operational variables such as water and solid split ratios, overflow and underflow densities, apex and spigot flowrates can easily be incorporated as the input parameters in the prediction of d50c. The ANN yields high correlation of data, hence it can be used in automatic control and multiphase operations of hydrocyclones.

An application of Kohonen's self-organizing map (SOM), learning-vector quantization (LVQ) algorithms, and commonly used backpropagation neural network (BPNN) to predict petrophysical properties obtained from well-log data are presented. A modular, artificial neural network (ANN) comprising a complex network made up from a number of subnetworks is introduced. In this approach, the SOM algorithm is applied first to classify the well-log data into a predefined number of classes, This gives an indication of the lithology in the well. The classes obtained from SOM are then appended back to the training input logs for the training of supervised LVQ. After training, LVQ can be used to classify any unknown input logs. A set of BPNN that corresponds to different classes is then trained. Once the network is trained, it is then used as the classification and prediction model for subsequent input data. Results obtained from example studies using the proposed method have shown to be fast and accurate as compared to a single BPNN network.

A system based on coded infrared signal transmission for the estimation of position of mobile robots in a structured environment is reported. Particular emphasis is placed on the polar coordinate arrangement in which signals are sent from the transmitters situated at the corners of the boundaries of operation. A multisensor system, strategically situated onboard the robot, has been found to improve the accuracy of the position estimation substantially. The information detected by the sensors is suitably processed to calculate the central position of the robot geometrically. The algorithms for the position calculations and the operational strategy are presented. This system forms the basis for the coordination and cooperation philosophy of multiple mobile robots sharing the same environment and performing cooperative or competitive tasks.

A micro robot has been built and integrated into a structured infrared positioning sensing system. In the position sensing system the operation area is divided into sectors to allow multiple robots to operate simultaneously. The robot monitors its position whilst in motion and makes its own logical decisions. It can autonomously decide for the next action to execute. It is also able to avoid both dynamic or static obstacles by either sensing at possible collision or by anticipating obstacles from its position and position of others by means of acting on a set of given prioritised instructions. Errors in the system have been discussed and recommendations are made

This paper demonstrates an application of artificial neural network (ANN) for determination of underflow and overflow densities of hydrocyclone separators. The discussions are extended and further results are presented for the prediction of particle size distributions in the underflow and overflow streams. The fit of the experimental results against the predicted results are illustrated and a statistical analysis is made. It is shown that, once the history of the operations are known, the ANN proves to he a useful tool for predicting future separation efficiencies. This approach has a potential to eliminate the need for installation of expensive on-line instruments for density measurements and particle size analyses. This approach can be applied in similar situations in the mineral processing industry.

This paper reports the application of Kohonen's Self-Organizing Map (SOM) and Learning Vector Quantization (LVQ) algorithms, and the commonly used Back Propagation Neural Network (BPNN) to the prediction of petrophysical properties from well log data. Recently, the use of artificial neural networks (ANN) in the field of petrophysical properties prediction has received increasing attentions. In this paper, a modular ANN comprises of a complex network made up of a number of sub-networks is introduced. In this approach, the SOM algorithm is first applied to classify the well log data into a pre-defined number of classes. This gives an indication of the lithology of the given well. The LVQ algorithm is then applied to train the network under supervised learning. A set of BPNN which corresponds to different classes is then developed for the prediction of petrophysical properties. Once the network is trained if is then used as the classification and prediction model for subsequent input data. Results obtained from example studies using this proposed method have shown to be fast and accurate as compared to a single BPNN network.

The hydrocyclone parameter d5Oc is estimated by the application of Artificial Neural Network (ANN). Results indicate that ANN performs equally well as compared to statically derived analytical formulae. Results also show that ANN will also be convenient in the on-line control of hydrocyclones.

Validation has been used for the estimation of generalisation error of the backpropagation networks. The simplest way is to divide the available data into training and validation data sets. An approach using the self-organising map is proposed for the selection of the training and validation data sets. The results obtained from this study has shown that the proposed method provides a quick and reliable selection criteria and the overall training time is also reduced by applying the split-sample early stopping approach

This paper reports the application of Kohonen's self-organising map (SOM) network to the classification of lithology from well log data. The well log data are classified into nodes according to a pre-defined grid arrangement. The learning vector quantization (LVQ) algorithm is then applied to train the network under supervised learning. After the network is trained, it is used as the classification model for subsequent data. Results obtained from example studies using this proposed method have shown to be fast and accurate